Can networking get real?

If you looked recently at the companies started by venture capitalists and touted as the "next Cisco" or "next Google," you know the meaning of the term "bad ending." Start-ups don't do well generally, and technology start-ups seem to be faring particularly badly. Recently, Sequoia Capital, one of the major Silicon Valley venture capital firms, pointedly told its portfolio companies "Get real, or get out." So, can start-ups get real, particularly in networking?

For openers, it's kind of ironic that a venture capital firm would say, "Get real!" This from the guys who spawned the tech bubble, financed tech companies, financed competitive local exchange carriers to buy from them, then called the result an industry. Truth be told, nobody in the world is as responsible for the "flip (sell the company) and walk" mind-set as the venture capitalists; and that's what has created the reality problem. I've listened to many venture capitalists say, "We don't want to be in that business; success takes too long."

Even before the current economic crisis, it was clear that networking was facing some major systemic problems. In the enterprise space, networking has received less than its average share of IT budgets every single year since 1991, and the numbers got worse in the last seven quarters. In the service-provider space, we're throwing money at companies whose business model consumes capacity rather than creates it, and whose revenue model depends totally on the shrinking pool of advertising.

Now we have our economic mess, and getting real isn't what got us into it. You can see from a comparison of government numbers that we've tried to grow "wealth" faster than "value," in the tech bubble and today. The charts of wealth vs. GDP in the Nasdaq bubble and the current crisis are strikingly similar. So, it's likely that regulation will try to rein in the excesses of the financial industry, and maybe even of the venture capitalists. What will we in the industry do? One thing might be to assume that a "real" business is one that can make a profit.

The time for running scams is over. We need to resolve the real issues in networking. Start with cost. Worldwide, service providers report 4.3 times as many support incidents with IP networks as with old-time TDM or frame relay. Today we have no greater spending on improving service and network operations, as a percentage of total infrastructure, than we did before. Until NEC bought NetCracker, no major vendor even had an operations-support-system product. Instead of figuring out how to do peer-to-peer to share files, we need to look at whether it could optimize operations. The cost of running a service-provider network is greater than the cost of building it, even discounting customer care. Standards groups are poking around at the problem but won't get to a solution for years, if at all. Why not entrepreneurs?

In the enterprise, we are getting all of the progress toward application-aware networks from the application side. You can get a great IBM or Microsoft pitch on service-oriented architectures, mashups, network middleware, and so forth. How about from network vendors? I asked a survey base of nearly 300 enterprises about network-vendor support for SOA and mashups, and what we heard was that nobody had ever had an organized presentation of the network impact of either -- from their network vendor. A quarter had heard it from IBM or Microsoft. Are SOA and mashups transparent to the network? Hardly; they could have a major impact on traffic and connectivity policies, and on capacity requirements. A network start-up to create an application network instead of a LAN or a WAN might be just the trick for the enterprise problem. Any takers?

I listened earlier in October to the CEO of a service-provider equipment company tell analysts that the current crisis would be short-lived, to which a Wall Street type remarked that short-lived meant "over by this afternoon." That puts finance and networking into perspective. Our buyers have three- to seven-year planning cycles, and we want it all done by the afternoon. In any event, it's pretty clear that the current crisis won't be short-lived, and that during its impact both enterprise and service-provider buyers are going to be looking for real answers, not venture-capitalist mumbo-jumbo.

America is largely responsible for the architectures that built the world's networks, just like our financial policies are responsible for the financial shenanigans that now have become the world's problem. We are at risk of surrendering both our financial leadership and our networking leadership now, and the networking part is threatened by our inability to move up from pushing bits around to focusing on what brings value to users. It's not the bit, it's the application or the service. For a decade, computing and software companies have eroded the network's place in the service layer by assuming that services are only on the network and never in it. If our Silicon Valley venture capitalists want start-up reality, the service layer for providers or the application layer for the enterprise is where they need to find it. The good news is that both are in one place. We talked "convergence" at the network layer, and we need an architecture for the service layer that's as easy to explain and as universally accepted.

My model says that the world economy is going to take a terrible hit, and that financial industry excess is the root cause. Let's admit that the attitudes of Wall Street are no strangers on Tech Street, and step out of our past to innovate again, where it counts. Who knows, real success might be even nicer than a bubble.

Join the Network World communities on Facebook and LinkedIn to comment on topics that are top of mind.

Copyright © 2008 IDG Communications, Inc.

SD-WAN buyers guide: Key questions to ask vendors (and yourself)